Summary
. US chipmaker taking precautions amid uncertainty over Beijing’s shipment approvals – sources
. China has requested tech firms temporarily pause H200 orders – sources
. Regulators still deciding how many domestic chips must accompany each H200 purchase – sources
BEIJING/SHANGHAI, Jan 8 (Reuters) – Nvidia (NVDA.O) is requiring Chinese customers to pay in full upfront for its H200 artificial intelligence chips, a precaution in response to ongoing uncertainty over Beijing’s approval of the shipments, according to two people familiar with the matter.
The U.S. chipmaker has implemented unusually strict terms, requiring full payment for orders without options for cancellation, refunds, or configuration changes once placed, the sources said.
In certain cases, clients may use commercial insurance or asset collateral instead of cash payment, one source added.
Previously, Nvidia’s standard terms for Chinese customers included advance payment requirements, but customers could sometimes pay a deposit rather than the full amount upfront, the source said. For the H200, however, the company is enforcing these rules strictly due to unclear regulatory approval in China, the source added.
Both sources spoke on condition of anonymity because the information is not public. The tightened policy has not been reported before. At the time of publication, Nvidia and China’s industry ministry had not responded to requests for comment.
Chinese technology companies have placed orders for more than 2 million H200 chips, priced at approximately $27,000 each, Reuters reported last month. This demand exceeds Nvidia’s inventory of 700,000 units.
While Chinese chipmakers such as Huawei have developed AI processors including the Ascend 910C, their performance still falls short of Nvidia’s H200 for large-scale training of advanced AI models.
China is expected to approve some H200 imports as early as this quarter, Bloomberg reported Thursday. Officials are preparing to allow purchases for certain commercial uses, while restricting military, sensitive government, critical infrastructure, and state-owned enterprises due to security concerns, the report said.
In recent days, Beijing has asked some Chinese tech firms to temporarily pause H200 orders while regulators determine how many domestic chips must accompany each H200 purchase, the second source said. The Information first reported the pause on Wednesday.
Nvidia CEO Jensen Huang said Tuesday that customer demand for H200 chips was “quite high” and that the company has “fired up our supply chain” to increase production. He added that he does not expect a formal announcement from China’s government, but “if the purchase orders come, it’s because they’re able to place purchase orders.”
BALANCING ACT
The strict upfront payment rules highlight the delicate balance Nvidia must maintain as it seeks to capitalize on growing Chinese demand while navigating regulatory uncertainty in both countries.
The Biden administration previously banned advanced AI chip exports to China, but President Donald Trump reversed that policy last month, permitting H200 sales with a 25% fee paid to the U.S. government.
Nvidia has faced setbacks in the past. Last year, it recorded a $5.5 billion inventory write-down after the Trump administration abruptly blocked sales of the H20 chip to China, previously its most powerful product in that market. While the U.S. has reversed the ban, China has since prohibited H20 shipments.
The H200’s payment structure shifts financial risk to customers, who must commit funds without certainty that Beijing will approve imports or that they can deploy the chips as intended.
Chinese internet giants, including ByteDance, see the H200 as a major upgrade over current offerings. Nvidia’s H200, the company’s second-most powerful chip, delivers roughly six times the performance of the now-blocked H20 chip that was specifically designed for China.
Nvidia intends to fulfill initial orders from existing stock, with the first batch of H200 chips expected before the Lunar New Year holiday in mid-February, Reuters reported last month.
The company has engaged contract chipmaker Taiwan Semiconductor Manufacturing Co (TSMC) to increase H200 production to meet Chinese demand, with additional manufacturing slated to begin in the second quarter of 2026, Reuters reported last week.
Expanding capacity remains challenging as Nvidia transitions from its current top chip, Blackwell, to the more advanced Rubin, while competing with companies such as Alphabet’s Google for advanced chipmaking capacity at TSMC.